Resolving Cost Management and Key Pitfalls of Agile Software Development - Daniel Saroff
Download as PPTX, PDF0 likes238 views
The document discusses the challenges of cost management in agile development, emphasizing the importance of measurement and realistic estimation for effective budget control and performance monitoring. It introduces Function Point Analysis (FPA) as a quantitative method for evaluating software functionality, offering consistent metrics for benchmarking and improving team performance. The analysis aims to enhance visibility, predictability, and management of software development costs, ultimately leading to better outcomes in agile environments.
Resolving Cost Management and Key Pitfalls of Agile Software Development - Daniel Saroff
- 1. Resolving Key Cost Management Pitfalls for Agile Development Daniel Saroff Vice President, Consulting IDC dsaroff@idc.com +1.508.935.4462
- 2. 2 © IDC | IDC | Technology Research,Advisory, and Consulting Global provider of market intelligence, insights, and data for nearly 60 years • Digitally-Capable Infrastructure • Leadership & Governance Strategies • Data, AI & Intelligence • The Trusted & Secure Enterprise • Systems of Experience: Customers & Employees • Metrics & KPIs • Innovation & Cost Management Future IT Industry Transformation • Executive Peer-to-Peer Connections • Brand Management Services • Executive Coaching and Leadership Development • Custom Development Services • Women in IT Coaching & Development Programs • CIO Board Readiness Advisory Services Use Case Centric Innovation • Securing & Documenting Cost Mgt Effectiveness • IT Deal Benchmarking • Tech Evaluation, RFP Support & Vendor Selection • Innovative technology acquisition • Audit Review and Defense • Contract Language Vendor Sourcing & Management Robotics 5G & Mobile Big Data / Analytics Blockchain Cloud AR/VR IoT CyberSecurity & Digital Trust 3D Printing AI Oil & Gas Utilities Mining Manufacturing Retail Hospitality & Dining Health Life Sciences Financial Services Insurance Government Smart Cities Education • Agile Improvement • Benchmarking • Cloud Cost Management • Cloud Readiness • IT Service & Vendor Cost Management • Maturity Assessment • Optimization • Sourcing • Strategy Assessment • Staffing Assessment IDC Metri Consulting
- 3. 3 © IDC | You can’t manage*, what you can’t measure *This include costs
- 4. 4 © IDC | For agile development, to understand: • The measurement challenge • Predictability, cost accountability, control, and visibility • Managing budgets • Increase delivery speed and product quality • Assessing agile team performance Goals
- 5. 5 © IDC | Software is costly • Software cost - US$542b / year • COTS – US$234b • Custom/bespoke – US$132.2b • Middleware – US$141b • Average cost overrun • 30% Hasn’t improved with shift to Agile from Waterfall. Why?
- 6. 6 © IDC | The promise and reality of agile Promise • Higher quality • Minimum viable product • Risk reduction • Speed • Transparency • Predictability • Lower costs • Flexibility • Customer satisfaction Reality • Cost blackhole • Variable quality • Difficult to manage • Challenges in visibility • Predictability challenges • Qualitative metrics • Challenge to compare across teams or projects
- 7. 7 © IDC | Challenges organizations experience Challenges in Monitoring and Reporting the Performance Q. Which of the following challenges, if any, is your organization facing with respect to monitoring and reporting the performance of your software development and DevOps teams?* *IDC survey
- 8. 8 © IDC | Agile does work ~3x ~1/2
- 9. 9 © IDC | What drives development cost Complexity Duration Effort Efficiency Productivity Delay Quality Change Turn-over Reliability Over / Under Estimation Resourcing Stability Security Maintenance …makes estimation seem hard
- 10. 10 © IDC | Realistic estimation is key for cost estimation Optimism Additional Costs Low Realistic High 100% 0% Pessimism Human Judgement
- 11. 11 © IDC | Blame it on story points • Story points are qualitative / subjective measures of effort • Teams estimate the effort an item requires compared to a reference story • Useful for development teams. Challenging for cost estimation and benchmarking • People find it difficult to estimate absolute things, but can easily estimate relatives • Story points measure effort not size. Effort is variable Cost estimation requires a consistent measure
- 12. 12 © IDC | Loss of value over time – burning money Sprint 1 Sprints 2…N MVP Production 90% 70% 50% 20%
- 13. 13 © IDC | Enough of the problem… …let’s get to the answer!
- 14. 14 © IDC | Source: Galorath
- 15. 15 © IDC | Function Point Analysis (FPA)* measures application functionality delivery. • International standards (e.g., OMG, CISQ, Nesma, IFPUG, COSMIC) • Standardized, weighted criteria that measures functional size • Normalized • Derives the Function Point Index (size and complexity of the software) • Can be applied across teams and projects and between companies • Quantitative • Doesn’t impact the work-flow of agile • Counting can be automated (and can be computed manually) *Allan Albrecht, 1979, “Measuring Application Development Productivity”
- 16. 16 © IDC | Function points – the details • Function Points measure standardized output • Enhanced Function Points (EFP) quantifies functionality added or changed • Automated Function Points (AFP) is FPA and EFP count automated by tools • Manual Function Points (mFPA) is manually FP counting The Simple Formulae • Productivity = Hours per EFP • Cost Efficiency = Costs per EFP • Speed = EFP per month or sprint • Project Quality = Defects per EFP The Simple Formulae • Hours = Team bandwidth • Cost = Salaries, technology, tooling • Speed = Productivity over time • Defects = Poor coding or CVEs
- 17. 17 © IDC | Software cost estimation maturity model Source: Galorath Agile Model Qualitative Function Pts. Quantitative Cost Estimate = Backlog or Requirements x Cost Cost Forecasting = (Backlog – Productivity) x Cost
- 18. 18 © IDC | Measuring to improve and benchmark Team Metrics Product Metrics High Performing Team Industry Average Low Performing Team -80% -60% -40% -20% 0% 20% 40% 60% 80% 100%
- 20. 20 © IDC | Usage • Multiple development teams • Anecdotally, some better than others • Benchmark performance and compare with FPA • Identify high performers, assess success drivers • Share practices from high performing teams to others
- 21. 21 © IDC | Usage • Low productivity • Updates slow • Security/CVE issues • FPA identifies refactoring churn (low maintainability / robustness) • Understanding this, you can focus on refactoring and improving maintainability and changeability • Better productivity, faster time to value, better costs mgt
- 22. 22 © IDC | Benefits Management Benefits Team Benefits Consistent reporting Faster development Visibility Improved requirements management Comparability Proper resourcing Predictability Predictability Cost forecasting Reduced stress Cost management Focus Project cost estimation Improved quality Business benefit analysis Doesn’t impact the agile methodology
- 23. 23 © IDC | • Because these metrics are based on an international standard (EFP), they can be used to benchmark teams and against industry metrics • Following trends for each team is feasible allowing for course corrections and targets to address performance • Additionally, these metrics may be included in third-party contracts to ensure supplier performance. This promotes suppliers to employ better developers to meet the quality metrics of the contract Benefits